european urology 55 (2009)

Transcription

1 european urology 55 (2009) available at journal homepage: Bladder Cancer Potential Impact of Postoperative Early Complications on the Timing of Adjuvant Chemotherapy in Patients Undergoing Radical Cystectomy: A High-Volume Tertiary Cancer Center Experience S. Machele Donat a, *, Ahmad Shabsigh a, Caroline Savage a, Angel M. Cronin c, Bernard H. Bochner a, Guido Dalbagni a, Harry W. Herr a, Matthew I. Milowsky b a Department of Urology, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, United States b Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, United States c Department of Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, United States Article info Article history: Accepted July 2, 2008 Published online ahead of print on July 14, 2008 Keywords: Bladder cancer Radical cystectomy Postoperative complications Adjuvant chemotherapy Abstract Background: Perioperative cisplatin combination chemotherapy is associated with a survival benefit in patients with invasive bladder cancer (BCa). However, in a recent report from the National Cancer Database (NCDB), only 11.6% of stage III BCa patients received perioperative chemotherapy, the majority in the adjuvant setting. Objective: We explore the impact of postoperative complications on the timing of adjuvant chemotherapy. Design, setting, and participants: An independent review board approved the review of 1142 consecutive radical cystectomies (RC), and data from these cases were entered into a prospective complication database ( ) which was utilized and retrospectively reviewed for accuracy at a single, academic, tertiary cancer center. Interventions: All patients underwent RC/urinary diversion by highvolume, fellowship-trained, urologic oncologists. Measurements: All complications within 90 d of surgery were defined and graded using a five-grade modification of the original Clavien system utilized at Memorial Sloan-Kettering Cancer Center and stratified into 11 categories. Grade 2 5 complications typically prohibit starting adjuvant chemotherapy. Univariate and multivariable logistic regression were used to evaluate variables associated with complications. Results and limitations: Overall, 64% (735 of 1142 patients) experienced one or more complications, of which 83% (611 of 735) were grade 2 5. * Corresponding author. Memorial Sloan-Kettering Cancer Center, Department of Urology, 1275 York Ave, Rm 519 Kimmel Bldg., New York, NY 10021, United States. Tel ; Fax: address: donats@mskcc.org (S.M. Donat) /$ see back matter # 2008 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi: /j.eururo

2 178 european urology 55 (2009) Furthermore, 57% of grade 2 5 complications (347 of 611) occurred between discharge and 90 d, 38% (233 of 611) within 6 wk, and 19% (114 of 611) between 6 wk and 12 wk, the general time frame for adjuvant chemotherapy. Overall, 26% (298 of 1142 patients) required readmission. Surgical morbidity at a high-volume tertiary cancer center may not reflect the case mix or surgical experience seen in the community setting. Conclusion: This series demonstrates that 30% of patients (347 of 1142) undergoing RC may not have been able to receive adjuvant chemotherapy due to postoperative complications. This information should be taken into consideration when planning multimodal therapy and further supports the use of perioperative chemotherapy in the neoadjuvant setting. # 2008 European Association of Urology. Published by Elsevier B.V. All rights reserved. 1. Introduction Combined modality treatment with neoadjuvant cisplatin-based chemotherapy and radical cystectomy (RC) has been shown to provide a significant survival benefit for patients with muscle-invasive bladder cancer [1 5], yet recent analyses from the Surveillance, Epidemiology, and End Results (SEER) Medicare Database [6] and National Cancer Database (NCDB) [7] indicate that chemotherapy is underutilized, with only 11 12% of eligible patients receiving perioperative chemotherapy and the majority receiving treatment in the adjuvant setting (10.4% adjuvant vs 1.2% neoadjuvant). Relevant factors may be a perceived risk of increased surgical complications after chemotherapy, fear of disease progression by delaying cystectomy, inadequate renal function or other patient-related comorbidities, and the belief that treatment based on pathologic criteria will permit the selection of high-risk patients who are most likely to benefit, thus avoiding unnecessary treatment and potential complications in patients with low-risk disease. Much of the debate on neoadjuvant versus adjuvant chemotherapy has revolved around potential complications related to chemotherapy and survival outcomes, with little attention placed on the potential risk of postsurgical complications on the ability to deliver adequate chemotherapy in a timely fashion [8 11]. Contemporary, large, single institution series report the risk of perioperative complications with RC in the range of 25 57%, inhospital mortality of 3%, and reoperative rates in the range of % [12]. This study explores the potential impact of perioperative complications on the ability to deliver adjuvant chemotherapy and the implications for multimodality treatment planning in patients with locally advanced bladder cancer undergoing RC. 2. Materials and methods An independent review board approved a review of 1142 consecutive RC patients, and data from these cases were entered into a prospective complication database ( ) which was utilized and retrospectively reviewed for accuracy. Since all patients had the potential need for adjuvant chemotherapy following RC, all stages were included in the data set. All complications within 90-d of surgery were defined, categorized, and classified by an established, five-grade modification of the original Clavien system utilized at Memorial Sloan-Kettering Cancer Center [13]. The grading system for complications utilized is the following: grade 1 complications required oral medication/bedside care; grade 2 complications required intravenous therapy, hyperalimentation, enteral feedings, or transfusion; grade 3 complications required intubation, interventional radiology, or re-operative intervention; grade 4 complications required organ resection or resulted in a chronic disability; and grade 5 complications resulted in death. Complications were defined and then placed into 11 categories: bleeding, surgical injury, gastrointestinal [GI], thromboembolic, wound, infection, genitourinary [GU], cardiac, pulmonary, neurologic, and miscellaneous. Grade 3 5 complications were considered significant due to the level of patient impact. Patients with grade 2 5 complications have the potential for being excluded from receiving adjuvant chemotherapy. The optimal time period for initiation of adjuvant chemotherapy was defined as 6 12 wk following surgery Statistical methods Univariate and multivariable logistic regression were used to evaluate variables associated with experiencing a complication. Variables analyzed were gender, age at surgery, BMI (body mass index; continuous and categorized as >30 kg/m 2 and 30 kg/m 2 ), prior abdominal surgery, preoperative chemotherapy, preoperative radiotherapy, operating-room time, estimated blood loss (EBL), total number of packs of red blood cells (PRBCs) received (>4 vs 4), total number of packs of fresh frozen plasma (FFPs) received (>4 vs 4), type of urinary diversion (continent vs conduit), American Society of Anesthesiologists (ASA) score (2 vs >2), Charlson-Romano

3 european urology 55 (2009) Table 1 Summary of all complications within 90 d of surgery by category Category type Complications included by category No. patients * (%) Gastrointestinal Ileus, small bowel obstruction, emesis, peptic ulcer, anastomotic bowel leak, 335 (29%) enterocutaneous fistula, ascites, GI bleed, diarrhea, c. difficile Infection Fever of unknown orgin, pelvic/retroperitoneal abscess, urinary tract infection, 282 (25%) pyelonephritis, cellulitis other than incisional, peritonitis, diverticulitis, cholecystitis, sepsis Wound Dehiscence, wound seroma, wound infection, cellulitis 168 (15%) Cardiac Myocardial infarction (MI), Arrhythmia, Congestive heart failure, hypotension, 131 (11%) hypertension, Ischemia without MI, angina Genitourinary Acute renal failure, hydro, ureteral stricture, urinary leak (anastomosis or pouch), 120 (11%) urinary fistula to bowel or skin, urinary retention, bladder neck contracture, urinary ascites, parastomal hernia, stomal stenosis, venous congestion/ischemia stoma Pulmonary Atelectasis, pneumonia, ARDS, dyspnea, pneumothorax, pleural effusion, empyema 98 (9%) Bleeding Anemia requiring transfusion, significant (1 l) intraoperative or postoperative 104 (9%) hemorrhage, flank hematoma, wound hematoma, scrotal hematoma, disseminated intravascular coagulopathy Thromboembolic Deep venous thrombosis, pulmonary embolus, superficial phlebitis, subclavian vein 93 (8%) thrombosis Neurologic Nerve palsy, paralysis, loss of consciousness, agitation, delirium, cerebrovascular 56 (5%) accident, vertigo Miscellaneous Psych illness, tendonitis, dermatitis, acidosis, thrombocytopenia without bleeding, foot ulcer, 38 (3%) lymphocele, decubitus ulcer Surgical Incisional hernia, vascular injury, retained drain, rectal injury, obturator nerve injury, enterotomy 8 (1%) * Note: Study included 1142 patients, and patients may be represented more than once since they may have experienced more than one complication. score (2 vs <2), preoperative creatinine level (<1.4 vs 1.4), and organ-confined disease. All of these variables were included in the multivariable analysis, with the exception of total PRBCs and number of FFPs received, because these are highly correlated with EBL. Although Charlson-Romano and ASA scores are both indicators of comorbid conditions, in our cohort they were not significantly collinear (for example, only 57% of patients with Charlson-Romano score 2 had an ASA score >2), and both variables were therefore included in multivariable analyses. BMI was entered as continuous in the multivariable model. Separate analyses were performed for the outcome of experiencing a grade 2 or higher complication, and experiencing a grade 3 or higher complication. Statistical analyses were conducted using Stata 9.0 (StataCorp, College Station, TX, USA). 3. Results Overall, 64% of patients (735 of 1142) experienced one or more complications with a total of 1637 complications occurring in 1142 patients within a 90-d period from the date of surgery. Of those, 83% (611 of 735 patients) had complications classified as grade 2 5, the majority, 57% (347 of 611) of which occurred between discharge and 90 d postoperatively. Of these, 38% (233 of 611 patients) experienced the complication within 6 wk of discharge, and 19% (114 of 611) occurred between 6 wk and 12 wk after surgery, the optimal time for adjuvant chemotherapy. Therefore, up to 30% of patients (347 of 1142) in the series experienced a grade 2 5 complication after discharge in a time period that would potentially interfere with the administration of adjuvant chemotherapy. The median length of stay (LOS) for the series was 9 d overall (interquartile range [IQR]: 8 12). In addition, hospital readmission for surgeryrelated complications was required in 26% (298 of 1142 patients) within 90 d of surgery, and 34% (382 of 1142 patients) required emergency room visits. Only 5% (58 of 1142 patients) required an intensive care unit (ICU) admission postoperatively within 90 d of surgery. Two percent (25 of 1142 patients) required a second operative procedure during their hospitalization, and another 1% (13 of 1142 patients) required a second operative procedure after discharge within 90 d of surgery. Angiointerventional procedures were required in 6% (67 of 1142 patients) during the initial hospitalization and in another 5% (60 of 1142 patients), following discharge within 90 d of surgery. In-hospital mortality was only 0.9% (10 of 1142 patients) with an overall 30-d mortality of 1.5% (17 of 1142 patients), and a 90-d mortality of 2.7% (31 of 1142 patients). Overall, 1.2% (14 of 1142 patients) died of an acute cardiopulmonary event (myocardial infarction [MI]/pulmonary embolus [PE]) within 30 d of surgery, accounting for 82% (14 of 17) of perioperative deaths, with over half (57%, 8 of 14) of those occurring after discharge from the hospital. Eight percent of patients (93 of 1142) experienced a thromboembolic event (PE and/or deep venous thrombosis [DVT]) within 90 d of surgery.

4 180 european urology 55 (2009) Table 2 Relationship of the most common complication categories to time from surgery including all grades * Rank by frequency of complication Time from surgery 0 90 d (n = 1637) 0 30 d (n = 1316) d (n = 221) d (n = 100) 1 GI (24%) GI (26%) Infection (33%) Infection (37%) 2 Infection (21%) Infection (17%) GI (17%) GI (20%) 3 Wound related (12%) Wound related (12%) GU (14%) GU (16%) 4 Cardiac (9%) Cardiac (10%) Wound related (9%) Wound related (8%) 5 GU (9%) Pulmonary (7%) DVT/PE (7%) Bleeding (5%) GI: gastrointestinal complication; GU: genitourinary complication; DVT: deep venous thrombosis; PE: pulmonary edema. * Patients who experienced multiple complications of the same category are counted more than once. Table 3 Total number of postsurgical complications occurring at each grade, by patient and by complication only the highest grade complication was counted for each patient Grade By patient (%) By complication No complication 407 (36%) Grade (11%) 429 (26%) Grade (40%) 1009 (62%) Grade (12%) 177 (11%) Grade 4 2 (0%) 3 (0.2%) Grade 5 (death) 19 (2%) 19 (1%) Total The overall frequency of the types of complications (including all grades) that occurred within 90 d of surgery is shown in Table 1. GI, infectious, woundrelated, cardiac, and GU complications were the five most common categories of complications experienced within the first 90 d after RC, all occurring in >10% of patients. However, when stratified by 30-d time periods postoperatively, infection-related complications became more prominent, GI complications persisted, and GU complications such as ureteral strictures and hydronephrosis increased over time as outlined in Table 2. As outlined in Table 1, wound infections were recorded in the wound-related complication category and pneumonia was recorded in the pulmonary category; therefore the infection category reflects mostly intra-abdominal abscesses and sepsis in the early (0 30 d) postoperative period and pyelonephritis and/or pouchitis at 30 d and beyond. The summary of complications by highest grade experienced is summarized in Table 3. Since grade 2 5 complications by definition were more likely to interfere with the administration of chemotherapy, we analyzed the type and frequency of their occurrence over time, stratifying them by 30-d postoperative time periods as outlined in Table 4.The table demonstrates how the type and frequency of Table 4 Summary of all grade 2 5 postsurgical complication experienced within 90 d of surgery, in order of frequency and stratified by 30-d time periods * All complications (grades 2 5) 0 90 d after RC 0 30 d after RC d after RC days after RC Complication category n = 611 (%) Complication category n = 547 (%) Complication category n = 97 (%) Complication category n = 54 (%) GI 242 (40%) GI 214 (39%) Infection 53 (55%) Infection 28 (52%) Infection 240 (39%) Infection 175 (32%) GU 22 (23%) GI 16 (30%) Cardiac 115 (19%) Cardiac 103 (19%) GI 19 20%) GU 8 (15%) Bleeding 99 (16%) Bleeding 90 (16%) DVT/PE 14 (14%) Wound 6 (11%) DVT/PE 92 (15%) DVT/PE 76 (14%) Wound 12 (12%) Bleeding 4 (7%) Wound 88 (14%) Wound 71 (13%) Cardiac 12 (12%) DVT/PE 3 (6%) GU 85 (14%) Pulmonary 63 (12%) Pulmonary 5 (5%) Pulmonary 2 (4%) Pulmonary 68 (11%) GU 59 (11%) Neurologic 5 (5%) Neurologic 2 (4%) Neurologic 29 (5%) Neurologic 21 (4%) Bleeding 5 (5%) Miscellaneous 2 (4%) Miscellaneous 13 (2%) Miscellaneous 9 (2%) Miscellaneous 2 (2%) Surgical 1 (2%) Surgical 7 (1%) Surgical 6 (1%) Surgical 0 (0%) Cardiac 1 (2%) n is the total number of patients; RC: radical cystectomy; GI: gastrointestinal complication; GU: genitourinary complication; DVT: deep venous thrombosis; PE: pulmonary edema. * Note that the percentages add up to more than 100%, since a patient could have had more than one complication.

5 european urology 55 (2009) Table 5 Summary of preoperative characteristics in 1142 patients undergoing radical cystectomy (RC) Patient characteristics Median (IQR) or frequency (%) No. patients missing data Age (yr) 68 (60 75) Gender (no. males) 862 (75%) Race (no. white) 1035 (91%) 1 BMI 27.1 ( ) 1 Very obese (BMI 30) 297 (26%) 1 Overweight (BMI 25) 781 (68%) 1 Charlson-Romano score (30%) 60 ASA score (43%) 3 Preoperative creatinine level abnormal (>1.4) 295 (26%) 8 Preoperative hydronephrosis 206 (18%) 16 Stent or nephrostomy placed 102 (9%) 23 Preoperative chemo 132 (12%) Preoperative radiation 40 (4%) Prior abdominal and/or retroperitoneal surgery 539 (47%) 1 Organ-confined disease at RC 657 (58%) IQR: interquartile range; BMI = body mass index, ASA = American Society of Anesthesiologists. complications change over time. For example, in the time period of d postoperatively, there were only four complication categories (infection, GI, GU, and wound-related complications) occurring in 10% or more of patients during that time period as opposed to eight categories (GI, infection, cardiac, bleeding, DVT/PE, wound, pulmonary, and GU) of complications within the first 30 d after surgery. Grade 3 5 complications occurred in 13% of patients (153 of 1142) within 90 d of surgery, with 12% (19 of 153 patients) resulting in death (grade 5). The most common grade 3 5 complications out of the 11 categories occurring in 10% or more of patients within 90 d of surgery were GU related (28%) or infection related (23%). However, when stratified by 30-d periods in addition to the GU- and infection-related complications, pulmonary, cardiac, and GI complications also occurred in 10% or more of patients, with the three most common complication categories at 30 d and beyond including GU (42%), infection (36%), and GI (18%). To define the level of patient comorbidity in our study population, the patient characteristics are outlined in Table 5. The median age of the study population was 68 yr. The majority of patients (68%; 781 of 1142) were overweight by World Health Organization (WHO) BMI classification, with 26% (297 of 1142 patients) having a BMI of 30. Forty- Table 6 Multivariable logistic regression to evaluate variables associated with complications grade 2 or higher and complications grade 3 or higher * Variables analyzed Outcomes Complication grade 2 5, n = 611 Complication grade 3 5, n = 153 Odds ratio (95% CI) p value Odds ratio (95% CI) p value Gender (male vs female) 0.79 ( ) ( ) 0.3 Age at surgery (5 yr) 1.09 ( ) ( ) 0.04 Body mass index (kg/m 2 ) 1.02 ( ) ( ) 0.6 Prior abdominal surgery (yes/no) 0.95 ( ) ( ) 0.03 Preoperative chemotherapy (yes/no) 1.02 ( ) ( ) Preoperative radiotherapy (yes/no) 2.16 ( ) ( ) 0.3 Operating-room time (100 min) 0.96 ( ) ( ) 0.13 Estimated blood loss (500 cm 3 ) 1.14 ( ) ( ) 0.04 Type of urinary diversion (continent vs conduit) 1.63 ( ) ( ) 0.7 ASA score (3 4 vs 1 2) 1.22 ( ) ( ) Charlson-Romano Score (2vs <2) 0.87 ( ) ( ) 1 Abnormal preoperative creatinine level >1.4 (yes/no) 1.04 ( ) ( ) 0.6 Organ-confined disease (yes/no) 1.26 ( ) ( ) 0.8 n is the number of events; ASA: American Society of Anesthesiologists. * 1066 patients are included in the multivariable analysis.

6 182 european urology 55 (2009) three percent (48 of 1142 patients) had an ASA score of 3 4, 40% had an age-adjusted Charlson-Romano morbidity index of >2, 23% had an age-adjusted Charlson-Romano morbidity index of 4, and 27% (303 of 1142 patients) had abnormal renal function (creatinine levels 1.4) preoperatively. The median operative time was 383 min with a median EBL of 1000 cm 3. Continent diversions were performed in 37% (418 of 1142 patients). A univariate analysis to evaluate predictors for significant (grade 3 5) complications following RC identified advancing age ( p = 0.02; odds ratio [OR] = 1.10 for each half-decade; 95% CI, ), EBL ( p = 0.013; OR = 1.12 for each 500 cm 3 ; 95% CI, ), transfusion >4 PRBCs ( p < ; OR = 3.22; 95% CI, ), transfusion >4 FFPs ( p < ; OR = 7.10; 95% CI, ), and ASA score >2 (p = 0.05; OR = 1.40; 95% CI, ) as significant predictive factors. Gender, BMI, prior abdominal/retroperitoneal surgery, preoperative chemotherapy/radiation, operating-room time, type of urinary diversion, Charlson-Romano score (2 vs < 2), renal function, and stage of disease were not significant predictors in univariate analysis ( p > 0.05, data not shown). In multivariable analysis, advancing age ( p = 0.04; OR = 1.12 for each half-decade; 95% CI, ), prior abdominal surgery ( p = 0.03; OR = 0.67; 95% CI, ), and EBL ( p = 0.04; OR = 1.11 for each 500 cm 3 ; 95% CI, ), were significant predictors of grade 3 5 postoperative complications (Table 6), while ASA score and preoperative chemotherapy had borderline results ( p = and 0.055, respectively), and operating-room time showed a marginal association ( p = 0.13). A multivariable analysis to predict grade 2 or higher complications identified the same significant predictors, except for prior abdominal surgery, which was no longer statistically significant ( p = 0.7; OR = 0.95; 95% CI, ). The type of urinary diversion ( p = 0.002) and preoperative radiation ( p = 0.04) were also significant predictors of grade 2 or higher complications, with ASA score again showing a trend such that those with an ASA score >2 were more likely to have a grade 2 or higher complication ( p = 0.15; Table 6). 4. Discussion Radical cystectomy is a morbid procedure even in the most experienced hands, with contemporary singleinstitution and population-based series reporting postoperative complications in the range of 25 57% [12]. The disparity in the incidence of complications between series is most likely a manifestation of the variety and accuracy of methodology utilized in reporting [13 16] and the lack of adjustment of case mix as suggested by Hollenbeck et al [17,18]. In the present series, where a standardized reporting methodology [13 15] is utilized and complications are defined, graded, and collected prospectively, the incidence of postoperative complications (grades 1 5) was 64%, with GI, infection, and wound-related complications the most common categories of complications seen overall (Tables 1 and 2). If we had chosen to only report the grade 3 5 complications which translate into what most series report as major complications [13 15], the incidence of postoperative complications in this series would have been only 13% (153 of 1142 patients), and the potential impact on adjuvant therapy would have been underestimated. This lack of systematic grading of complications in surgical series is contrary to what is seen in the medical oncology literature where the Common Terminology Criteria for Adverse Events (CTCAE) developed by the National Cancer Institute is the standard used to report adverse events associated with clinical trials [19]. This becomes especially important when planning combined modality treatments for bladder cancer where the patient population is older (median age of 68 yr in this series) with significant comorbidities, and therefore at greater risk for treatment-related complications [17 18]. Most clinicians would agree that combined modality treatment, with cisplatin-based combination chemotherapy and RC, provides a significant survival benefit for patients with locally advanced muscleinvasive bladder cancer [1 5]. Yet controversy remains as to whether it is better to receive neoadjuvant chemotherapy or adjuvant treatment based on pathologic findings, due to the inaccuracy of clinical staging and the potential toxic effects of chemotherapy on those who may not necessarily benefit from systemic therapy [8 11]. Although there is toxicity data on the commonly used chemotherapy regimens and patient-related factors limiting its use [20 23], including renal insufficiency [24,25], little prospective data exists concerning the effect of postsurgical complications on the ability to deliver at least three cycles of systemic chemotherapy in the postoperative setting. In one of the few randomized trials providing any data on perioperative surgical complications [11], the authors found that administering multiagent chemotherapy such as methotrexate vinblastine doxorubicin and cisplatin (M-VAC) in the postoperative setting is difficult, with only 77% of patients in the adjuvant arm receiving at least two cycles of therapy versus 97% in the neoadjuvant arm,

7 european urology 55 (2009) indicating that patients may be more likely to receive therapeutic chemotherapy (three cycles) if they are given in the neoadjuvant setting. In addition, the morbidity of combined modality treatment was most prominent in older patients with only 20% of those >70 yr of age able to complete the combined treatment in any order versus 7% in those <70 yr of age. In this series, we explore the potential impact of postoperative complications on the timing of adjuvant therapy by employing a standardized methodology for data collection and reporting [13 15] utilizing the Memorial Sloan-Kettering Cancer Center derivation of the Clavien system as outlined previously in the materials and methods section. The optimal timing for initiation of adjuvant chemotherapy was defined as 6 12 wk following surgery, therefore all complications up to 90 d after surgery were recorded and analyzed. Due to the fact that grade 2 complications by definition require intravenous therapy, hyperalimentation, enteral feedings, or transfusion have the potential to delay chemotherapy, the analysis included grades 2 5 (Tables 4 and 6) in addition to a separate accounting of grade 3 5 complications. In addition, to evaluate the evolving nature of complications, all complications were analyzed in blocks of 30 d as demonstrated in Table 4, lending some insight into the predominate types of complications seen over time. The most significant predictors of grade 3 5 complications in multivariable analyses were advancing age, prior abdominal surgery, higher EBL, and higher ASA score. Interestingly, when grade 2 complications were included in the analysis, the type of urinary diversion and prior radiation therapy also became significant predictors of complications. These findings indicate the importance of utilizing a grading system with defined complications rather than arbitrarily separating complications into major versus minor categories when assessing predictive variables and surgical outcomes, and may, in addition to the case mix [17,18], account for some of the disparity in surgical outcomes among RC series. This series confirms the high morbidity of RC in an elderly comorbid population with experienced surgeons in a high-volume institution. Overall, 64% of patients (735 of 1142) in our series experienced one or more complications within a 90-d period from the date of surgery, with 17% (124 of 735) of those suffering only a minor grade 1 complication requiring only oral medication or bedside care. Eighty-three percent (611 of 735 patients) suffered a grade 2 5 complication, the majority 57% (347 of 611) of which occurred between discharge and 90-d following surgery. Of these, 38% (233 of 611 patients) experienced the complication within 6 wk of discharge, and 19% (114 of 611) occurred between 6 wk and 12 wk after surgery, the optimal time for adjuvant chemotherapy. Therefore, up to 30% of patients (347 of 1142) in the series experienced a grade 2 5 complication after discharge in a time period with the potential to interfere with the administration of adjuvant therapy. Overall 26% (298 of 1142 patients) required hospital readmission for surgery-related complications, 1% (13 of 1142 patients) required a second surgical procedure after discharge, 5% (60 of 1142 patients) required an angiointerventional procedure after discharge, and 34% (382 of 1142 patients) required an emergency room visit within 90 d of surgery. The weaknesses of the study are that we may have under-reported some complications that were treated by the local physician and not captured in the database, although all efforts were made to do so. In addition, one could argue that some patients who suffered complications may have recovered in time to receive chemotherapy; however, this is unpredictable and varies according to individual patient comorbidities and age, again pointing out the importance of adjusting for case mix when assessing the impact of complications. We also did not assess the time-to-recovery for each type of complication because this data was not captured in the prospective database and would be impossible to reliably recover retrospectively. Of note, on the contrary, some grade 1 complications such as large superficial wound separations requiring local wound care and oral antibiotics, which might delay the start of chemotherapy due to the risk of infection or dehiscence, although recorded in the overall complications, were not included in the calculation for possible impact on chemotherapy. 5. Conclusion This series demonstrates that, when a comprehensive complication reporting system is utilized, up to 30% of patients undergoing RC may experience a perioperative complication which could preclude or delay their ability to receive adjuvant chemotherapy. This information should be taken into consideration when planning multimodal therapy and clinical trials for patients with invasive bladder cancer, and further supports the use of perioperative chemotherapy in the neoadjuvant setting. Furthermore, surgical complications have the potential to impact the timing and tolerability of chemotherapy and therefore should be included in

8 184 european urology 55 (2009) assessing the effectiveness of therapy in clinical trials. Author contributions: S. Machele Donat had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Donat, Shabsigh, Milowsky. Acquisition of data: Donat, Shabsigh. Analysis and interpretation of data: Donat, Shabsigh, Savage, Cronin. Drafting of the manuscript: Donat, Shabsigh, Milowsky. Critical revision of the manuscript for important intellectual content: Donat, Shabsigh, Milowsky, Bochner, Dalbagni, Herr. Statistical analysis: Savage, Cronin. Obtaining funding: None. Administrative, technical, or material support: None. Supervision: None. Other (specify): None. Financial disclosures: I certify that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None. Funding/Support and role of the sponsor: None. References [1] Grossman HB, Natale RB, Tangen CM, et al. Neoadjuvant chemotherapy plus cystectomy compared with cystectomy alone for locally advanced bladder cancer. N Engl J Med 2003;349: [2] Advanced Bladder Cancer Meta-analysis Collaboration. Neoadjuvant chemotherapy in invasive bladder cancer: a systematic review and meta-analysis. Lancet 2003;361: [3] Winquist I, Kirchner TS, Segal R, Chin J, Lukka H. Neoadjuvant chemotherapy for transitional cell carcinoma of the bladder: a systematic review and meta-analysis. J Urol 2004;171: [4] Herr HW, Dotan Z, Donat SM, Bajorin DF. Defining optimal therapy for muscle invasive bladder cancer. J Urol 2007;177: [5] Kassouf W, Spiess PE, Brown GA, et al. P0 stage at radical cystectomy for bladder cancer is associated with improved outcome independent of traditional clinical risk factors. Eur Urol 2007;52: [6] Schrag D, Mitra N, Xu F, et al. Cystectomy for muscleinvasive bladder cancer: patterns and outcomes of care in the Medicare population. Urology 2005;65: [7] David KA, Milowsky MI, Ritchey J, Carroll PR, Nanus DM. Low incidence of perioperative chemotherapy for stage III bladder cancer 1998 to 2003: a report from the national cancer data base. J Urol 2007;178: [8] Lehmann J, Retz M, Stockle M. Chemotherapy in the post- MVAC era: the case for adjuvant chemotherapy. World J Urol 2002;20: [9] Sonpavde G, Petylak DP. Perioperative chemotherapy for bladder cancer. Crit Rev Oncol/Hematol 2006;57: [10] Raghavan K, Quinn D, Skinner DG, Stein JP. Surgery and adjunctive chemotherapy for invasive bladder cancer. Surg Oncol 2002;11: [11] Millikan R, Dinney C, Swanson D, et al. Integrated therapy for locally advanced bladder cancer: final report of a randomized trial of cystectomy plus adjuvant M-VAC versus cystectomy with both preoperative and postoperative M-VAC. J Clin Oncol 2001;19: [12] Konety BR, Allareddy V, Herr H. Complications after radical cystectomy: analysis of population-based data. Urology 2006;68: [13] Jaques DP. Measuring morbidity. Ann Surg 2004;240: [14] Martin II RCG, Brennan MF, Jaques DP. Quality of complication reporting in the surgical literature. Ann Surg 2002;235: [15] Donat SM. Standards for surgical complication reporting in urologic oncology: time for a change. Urology 2007;69: [16] Miller DC, Filson CP, Wallner LP, Montie JE, Campbell DA, Wei JT. Comparing performance of morbidity and mortality conference and national surgical quality improvement program for detection of complications after urologic surgery. Urology 2006;68: [17] Hollenbeck BK, Miller DC, Taub D, et al. Identifying risk factors for potentially avoidable complications following radical cystectomy. J Urol 2005;174: [18] Hollenbeck BK, Miller DC, Taub DA, et al. The effects of adjusting for case mix on mortality and length of stay following radical cystectomy. J Urol 2006;176: [19] Trotti A, Colevas AD, Setser A, et al. CTCAE v. 3.0: development of a comprehensive grading system for adverse events of cancer treatment. Semin Radiat Oncol 2003; 13: [20] Loehrer Sr PJ, Einhorn LH, Elson PJ, et al. A randomized comparison of cisplatin alone or in combination with methotrexate, vinblastine, and doxorubicin in patients with metastatic urothelial carcinoma: a cooperative group study. J Clin Oncol 1992;10: [21] Logothetis CJ, Dexeus F, Sella A, et al. A prospective randomized trial comparing CISCA to MVAC chemotherapy in advanced metastatic urothelial tumors. J Clin Oncol 1990;8: [22] Saxman S, Propert K, Einhorn LH, et al. Long-term followup of a phase III intergroup study of cisplatin alone or in combination with methotrexate, vinblastine, and doxorubicin in patients with metastatic urothelial cancer: a cooperative group study. J Clin Oncol 1997;15: [23] Von der Maase H, Hansen SW, Roberts JT, et al. Gemcitabine and cisplatin versus methotrexate, vinblastine, doxorubicin, and cisplatin in advanced or metastatic bladder cancer: results of a large, randomized, multinational, multicenter, phase III study. J Clin Oncol 2000;18:

9 european urology 55 (2009) [24] Dash A, Galsky MD, Vickers AJ, et al. Impact of renal impairment on eligibility for adjuvant cisplatin-based chemotherapy in patients with urothelial carcinoma of the bladder. Cancer 2006;107: [25] Raj GV, Iasonos A, Herr H, Donat SM. Formulas calculating creatinine clearance are inadequate for determining eligibility for cisplatin based chemotherapy in bladder cancer. J Clin Oncol 2006;24: Editorial Comment on: Potential Impact of Postoperative Early Complications on the Timing of Adjuvant Chemotherapy in Patients Undergoing Radical Cystectomy: A High-Volume Tertiary Cancer Center Experience Robert S. Svatek Department of Urologic Oncology, The MD Anderson Cancer Center, Houston, Texas, United States Shahrokh F. Shariat Department of Urology, The University of Texas Southwestern, Medical Center, 5323 Harry Hines Boulevard, Dallas, TX , United States The optimal timing of administration of chemotherapy in relation to cystectomy is not known. Advantages to an adjuvant approach are the selective administration of chemotherapy to those patients based on precise pathologic staging, thereby limiting unnecessary treatment [1], avoiding delay in surgery, and alleviating patient anxiety [2]. Alternatively, advantages of a neoadjuvant approach include clear evidence of benefit from properly conducted trials, downstaging [3] and potential for improved resectability, and the ability to monitor treatment response [2]. In this edition of European Urology, Donat and colleagues [4] explore another important consideration in this debate: the impact of postcystectomy complications on the ability to deliver adjuvant chemotherapy. This study utilized a prospective, standardized reporting methodology to characterize the risk of perioperative morbidity following cystectomy [4]. This approach is superior to previous attempts to quantify perioperative morbidity based on retrospective chart review. The authors identified postcystectomy complications in 64% of patients and hospital readmissions in 26%. This study highlights the potential risk that postcystectomy complications may have on the ability to deliver adjuvant chemotherapy. We should mention several caveats, however, regarding the authors conclusions that 30% of patients undergoing cystectomy may not have been able to receive adjuvant chemotherapy and that their data support the use of neoadjuvant as opposed to adjuvant chemotherapy. First, some patients in this cohort received neoadjuvant chemotherapy or radiotherapy and this limits the application of these findings to patients typically considered to be candidates for adjuvant chemotherapy. Second, although 30% of patients may have had a delay in administration of adjuvant chemotherapy, not all 30% may have been excluded from receiving adjuvant chemotherapy based on postoperative complications. Grade 2 complications such as transfusion requirement or intravenous therapy would not preclude most patients from receiving adjuvant chemotherapy. Indeed in a trial in which patients receiving two cycles of neoadjuvant methotrexate, vinblastine, Adriamycin, cisplatin (MVAC) in addition to three postoperative cycles were compared to patients receiving five cycles of adjuvant MVAC without preoperative chemotherapy, <10% of patients (6 of 70) assigned to the immediate surgery arm were unable to receive adjuvant chemotherapy because of postoperative complications [5]. Finally, recovery from complications, a critical variable in determining the ability to administer chemotherapy, was not a component of the data collection. Although the efficacy of neoadjuvant chemotherapy is well established, the modest survival benefit has prompted some institutions to use a risk-adapted approach to neoadjuvant chemotherapy whereby selective administration is based on high-risk features such as lymphovascular invasion, ureteral obstruction, or clinical T3b disease [2,6]. This remarkable study by Donat and colleagues provides support for consideration of other factors, such as age or prior abdominal surgery, that may affect the ability to administer chemotherapy after cystectomy because of the increased risk for postoperative complications [4]. Furthermore, the standardized prospective recording of complications represents a substantial improvement in reporting methodology, and the staggering rates of postcystectomy morbidity in this series emphasize the need for improvement in perioperative care and careful preparation of patients undergoing cystectomy. References [1] Shariat SF, Palapattu GS, Karakiewicz PI, et al. Discrepancy between clinical and pathologic stage: impact on prognosis after radical cystectomy. Eur Urol 2007;51:

10 186 european urology 55 (2009) [2] Black PC, Brown GA, Grossman HB, Dinney CP. Neoadjuvant chemotherapy for bladder cancer. World J Urol 2006;24: [3] Kassouf W, Spiess PE, Brown GA, et al. P0 stage at radical cystectomy for bladder cancer is associated with improved outcome independent of traditional clinical risk factors. Eur Urol 2007;52: [4] Donat SM, Shabsigh A, Savage C, et al. Potential impact of postoperative early complications on the timing of adjuvant chemotherapy in patients undergoing radical cystectomy: a high-volume tertiary cancer center experience. Eur Urol 2009;55: [5] Millikan R, Dinney C, Swanson D, et al. Integrated therapy for locally advanced bladder cancer: final report of a randomized trial of cystectomy plus adjuvant M-VAC versus cystectomy with both preoperative and postoperative M-VAC. J Clin Oncol 2001;19: [6] Bartsch GC, Kuefer R, Gschwend JE, de Petriconi R, Hautmann RE, Volkmer BG. Hydronephrosis as a prognostic marker in bladder cancer in a cystectomy-only series. Eur Urol 2007;51: DOI: /j.eururo DOI of original article: /j.eururo